Automated identification of basalt spectra in Clementine lunar data

Abstract

The identification of fresh basalt spectra plays an important role in lunar stratigraphic studies; however, the process can be time consuming and labor intensive. Thus motivated, we developed an empirically derived algorithm for the automated identification of fresh basalt spectra from Clememtine UVVIS data. This algorithm has the following four parameters and limits: BC Ratio=3(R950−R900)/(R900−R750)<1.1, CD Delta=(R1000−R950)/R750−1.09(R950−R900)/R750>0.003 and <0.06, B Slope=(R900−R750)/(3R750)<−0.012, and Band Depth=(R750−R950)/(R750−R415)>0.1, where R750 represents the unnormalized reflectance of the 750nm Clementine band, and so on. Algorithm results were found to be accurate to within an error of 4.5% with respect to visual classification, though olivine spectra may be under-represented. Overall, fresh basalts identified by the algorithm are consistent with expectations and previous work in the Mare Humorum area, though accuracy in other areas has not yet been tested. Great potential exists in using this algorithm for identifying craters that have excavated basalts, estimating the thickness of mare and cryptomare deposits, and other applications.